Research on Indoor Positioning Method Based on Improved HS-AlexNet Model
Scene recognition is the key to achieving accurate and fast indoor positioning.Deep network has become a research central issue lately with its outstanding performance.This paper puts forward an advanced AlexNet network model combined with Harris feature detection,which guides the image processing of the original model according to the detected Harris characteristics,it reduces the randomness error and improves the generalization ability and robustness of the model.In addition,for the campus indoor scene positioning environment,the original structure of the AlexNet network model and the data augmentation method are improved,so that the positioning model can cope with the complex and variable positioning environment,and its accuracy and speed reach a high level.The method can be combined with the existing mainstream visual indoor positioning method to enhance the accurateness and speed of positioning system.
Scene recognition AlexNet network model Harris features Data augmentation Indoor positioning
Libiao Zhang Rui Zhao Yuqing Liu Xinyu Yang Shipeng Li
School of Information Science and Technology,Northeast Normal University,Changchun 130117,Jilin,China
国际会议
江苏镇江
英文
270-277
2019-09-20（万方平台首次上网日期，不代表论文的发表时间）